I'm interested in learning more about current limiting circuits. At the moment this is in the context of short circuit protection on outputs with relatively low power, anywhere from 25 to a few hundred mA at 5-12VDC (digital and analog signals, small display circuits, etc.)

Most of the simple, non op amp, circuits I've seen use the forward voltage of base emitter junction as a key element. Some of them seem to be highly dependent on the Hfe of the BJT, others less so, but most will give very different results when using different BJTs.

What I'd like to know is, how reliable and predictable is the Vbe junction voltage? I know the Hfe is highly variable, even within the same lot of the same part. Is the Vbe really unpredictable too? I'm not even sure how to spec things, because many parts only list a Vbe-sat, but no indication of what Vbe would be at the very low currents involved in these circuits.

I've been simulating a bunch of circuits and seeing how different transistors behave in them, but I'm not clear on how they'll behave in the real world with variable specs. Again, my question right now is specifically about Vbe in circuits like the first two below.

(P.S. Thanks to @crutschow, who designed the circuit that one of these is based on.)

 

What I'd like to know is, how reliable and predictable is the Vbe junction voltage? I know the Hfe is highly variable, even within the same lot of the same part. Is the Vbe really unpredictable too?

It's not predictable with any great precision; I usually figure Vbe, at modest collector currents, will be somewhere around 600-700 mV for most transistors. Plan on that much spread, perhaps a bit more.

 

Vbe is fairly consistent between transistors of the same part number at a given base current.
But for a given collector current, the base current will vary due to large differences in Hfe (Beta), and that will cause a variation in Vbe.
This variation is reduced at low collector currents.
For that reason, the MOSFET current-limit circuit will show less variation in current limit with different transistors of the same type.

Note that Vbe varies about -0.3%/°C so that can cause a significant variation in the current limit for a change in ambient temperature.

For best results with the MOSFET current-limit circuit you show, remove R3 and increase the value of R4 to 50k-100k to reduce the collector current.
R3 is only needed if the input voltage exceeds the maximum Vgs rating of the MOSFET.

 

It's not predictable with any great precision; I usually figure Vbe, at modest collector currents, will be somewhere around 600-700 mV for most transistors. Plan on that much spread, perhaps a bit more.

Cool. That's kind of what I initially expected, but I tried basically every pnp that came with LTspice just to appease my curiosity and there were a few crazy outliers that were way off. I might dig into their datasheets to see if I can make sense of them.

I'll do a little number crunching, but I imagine that amount of variation will be totally fine for the applications I'm picturing right now.

Also, I forgot to mention it in the first post, but thanks again for all your help on the modified current mirror circuit. You're the one who talked me through the calculations and application of that circuit. I've only used it once so far, but it's working beautifully!

 

Vbe is fairly consistent between transistors of the same part number at a given base current.
But for a given collector current, the base current will vary due to large differences in Hfe (Beta), and that will cause a variation in Vbe.
This variation is reduced at low collector currents.
For that reason, the MOSFET current-limit circuit will show less variation in current limit with different transistors of the same type.

Note that Vbe varies about -0.3%/°C so that can cause a significant variation in the current limit for a change in ambient temperature.

For best results with the MOSFET current-limit circuit you show, remove R3 and increase the value of R4 to 50k-100k to reduce the collector current.
R3 is only needed if the input voltage exceeds the maximum Vgs rating of the MOSFET.

Awesome! Thanks so much for the general explanations and for the specific advice on the MOSFET circuit.

I had wondered about the voltage divider on the gate, but hadn't thought to test without it yet. Everything I had in mind is 12V or less, so would work without it, but I'm glad I j know the intent so I don't abuse the circuit later with higher voltages!

I'll do some number crunching and some more simulations to try to better wrap my head around the collector->hfe->base relationships and variability. It's easier to see this stuff once you know what you're looking for, so thanks for the guidance.